364 Phillips. — On the Development of 
developed together. The 4-celled external branch is. how- 
ever, always present and always complete. Then again, while 
the number of cells in other filaments varies, though the 
number is usually three, the branch in question is invariably 
4-celled. In Plumaria , moreover, I have found the cells of 
other branches occasionally bud out peripheral cells, therein 
indicating their vegetative character (PL XVIII, Fig. 17). 
This never occurs in the more highly specialized cells of 
the true procarp. 
2. This 4-celled external branch is the only one which 
comports itself in its staining and other properties like the 
carpogonial branch in other Florideae. Treated with Hoff- 
mann’s blue, this branch stains more readily, and retains the 
stain more tenaciously, than any other branch in the group. 
Indeed, little distinction can be drawn between the behaviour 
of the other branches and those making up the involucre. 
All young cells, it is true, stain deeply with this re- agent, but 
the cells of the true carpogonial branch retain this capacity 
as long as they can be distinguished. Again, treated with 
glycerine, the cells of the true procarp swell up much more 
than those of other branches, a circumstance which may be 
seen even in Davis’s figures, where, however, the gelatinization 
of the walls is very general. The true trichogyne, too, is 
more gelatinous than the terminal hairs of other branches. 
In the latter 1 have often seen a bend with a kink, like that 
which appears in an india-rubber tube which is sharply bent. 
3. Thirdly, this 4-celled branch corresponds in form and 
juxtaposition to the carpogenous cell of the procarps of other 
Ceramiaceae. All Davis’s figures show faithfully a crescentic 
curvature in this branch, which permits it to be readily dis- 
tinguished. Other branches are straight, or only slightly 
curved ; this is so sharply bent as often to bring the carpo- 
gonium back to the mother-cell of the branch, and always 
sufficiently far to bring it into close contiguity with the 
carpogenous (auxiliary) cell derived from it. When Davis 
says that in Piilota the trichogyne (a term in which he 
includes both the trichogyne proper and the carpogonium 
